linux-kernel - Re: [PATCH v3] char: xillybus: Check endpoint type at probe time

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAMhUBjng1qc3iL8CeDZ+095xHr+LDy6ay9ojRfTho2ayTsO05w@mail.gmail.com>
Date:   Tue, 31 May 2022 11:50:18 +0800
From:   Zheyu Ma <zheyuma97@...il.com>
To:     Eli Billauer <eli.billauer@...il.com>
Cc:     arnd@...db.de, Greg KH <gregkh@...uxfoundation.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3] char: xillybus: Check endpoint type at probe time

On Mon, May 30, 2022 at 11:13 PM Eli Billauer <eli.billauer@...il.com> wrote:
>
> On 29/05/22 09:58, Zheyu Ma wrote:
> >   static int xillyusb_setup_base_eps(struct xillyusb_dev *xdev)
> >   {
> > +     int ret;
> > +     struct usb_endpoint_descriptor *in, *out;
> > +
> > +     ret = usb_find_common_endpoints(xdev->intf->cur_altsetting,&in,&out, NULL, NULL);
> > +     if (ret)
> > +             return ret;
> > +
> > +     if (in->bEndpointAddress != (IN_EP_NUM | USB_DIR_IN) ||
> > +             out->bEndpointAddress != (MSG_EP_NUM | USB_DIR_OUT))
> > +             return -EINVAL;
> > +
> >
> As far as I understand, this finds the first BULK endpoints in both
> directions, and verifies that their addresses are MSG_EP_NUM and
> IN_EP_NUM. Because both of these happen to equal 1, I suppose this
> indeed checks the right thing. But am I right that this won't work if
> either MSG_EP_NUM or IN_EP_NUM have a value that isn't 1? Not that I
> think that will ever happen, but still.

Indeed, the correctness of this code comes from the fact that both
MSG_EP_NUM and IN_EP_NUM are the first bulk endpoint, without such an
assumption this check does not hold.
I did this just for convenience, we can also use
xillyusb_check_endpoint() to check them.

> > +static int xillyusb_check_endpoint(struct xillyusb_dev *xdev, u8 addr)
> > +{
> > +     int i;
> > +     struct usb_host_interface *if_desc = xdev->intf->altsetting;
> > +
> > +     for (i = 0; i<  if_desc->desc.bNumEndpoints; i++) {
> > +             struct usb_endpoint_descriptor *ep =&if_desc->endpoint[i].desc;
> > +
> > +             if (ep->bEndpointAddress == addr&&  usb_endpoint_is_bulk_out(ep))
> > +                     return 0;
> > +     }
> > +
> > +     return -EINVAL;
> > +}
> > +
> >
> >
> Given that you've added this function, why isn't it used in
> xillyusb_setup_base_eps()?

It is feasible to use this function for checking in
xillyusb_setup_base_eps(), perhaps with some modifications. Anyway, I
will do so in the next version of the patch.

Thanks,
Zheyu Ma